Diameter trouble

[Abanks]

Hello all, i am a kinda new user and am having difficulty measuring diameters correctly. I can physically measure the diameters of the part i am achecking with calipers (which are calibrated) and get a good dimension. However when i measure the features with the CMM i get bad results. Seems to always show the diameters oversized thus throwing off my true position call outs. I am using a 2x40mm shank tip. Any help would be appreciated. I dont want to reject the parts when i know they are correct. Thanks all.

[John Riggins]

I am using a 2x40mm shank tip.

There is your problem.
Try the same diameters with a spherical tip.

[Abanks]

Ok thanks. only problem i have with that is, i am measuring sheet metal that is only 1mm in thickness. I used the cylindrical shape to make sure i had good surface contact. If i change to a ball probe wont i have to basically re-program my hits? Its a very complex part and that would be very time consuming. Thank you again for the info.

[John Riggins]

That does create problems.
One thing you could do (but this would take a lot of extra probing and therefore time):
This will require a small probe tip, maybe .5mm.
Probe auto-edge points at multiple points around the diameter.
Then use those points to construct your diameter.
This may give you better results.

[Schlag]

Not to beat a dead horse on this topic but....how many points are you taking ? Are you using LEAST SQ , MX IN , MIN CIR. Are your holes ROUND ( or close to )? can be allot of variables here. We have a shank probe that we use fairly often and they work great for certain applications, but you need to understand and anticipate their shortcummings and work around them.

[Abanks]

the feature itself shows using "Least SQ" not sure about the MX IN or MIN CIR, where do you adjust that? and does that change apply to the Position dimension also? yes holes are round with very good circularity. i thought it was the probe itself but it checks good in the Tip Qual on my test sphere. Its even worse with round slots!especially when they are 9.2mm x 9.6mm. i appreciate everyones feedback.
CIR13 =FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,IN,LEAST_SQR
THEO/<227,52.6,-2.469>,<0,-1,0>,4.5
ACTL/<227.261,52.6,-2.613>,<0,-1,0>,4.329
TARG/<226.969,52.6,-2.469>,<0,-1,0>
START ANG=0,END ANG=360
ANGLE VEC=<1,0,0>
DIRECTION=CCW
SHOW FEATURE PARAMETERS=NO
SHOW CONTACT PARAMETERS=YES
NUMHITS=3,DEPTH=0.5,PITCH=0
SAMPLE HITS=0,SPACER=0
AVOIDANCE MOVE=BOTH,DISTANCE=5
FIND HOLE=DISABLED,ONERROR=NO,READ POS=NO
SHOW HITS=NO

DIM 32= POSITION OF CIRCLE CIR13 UNITS=MM ,$
GRAPH=OFF TEXT=OFF MULT=10.00 OUTPUT=BOTH FIT TO DATUMS=OFF DEV PERPEN CENTERLINE=OFF DISPLAY=DIAMETER
AX NOMINAL +TOL -TOL BONUS MEAS DEV OUTTOL
X 227.000 227.261 0.261
Z -2.469 -2.613 -0.144
DF 4.500 0.200 0.000 0.000 4.329 -0.171 0.171 <--------
TP MMC 1.000 0.000 0.596 0.000 -----#---
END OF DIMENSION 32

[Abanks]

Use auto circle with sample hits. It will find the surface around the hole and measure down from there.

Thank you i might have to do that once i satisfy production.

[Liambo734]

Thank you i might have to do that once i satisfy production.

Use more than 3 hits, it's not enough, dead horse but you can't measure circularity with 3 hits, it will always look good, use 11 or 15 etc or 125 if you can sacrifice the time. Change the calculation to MAX_INSCR. If you use sample hits turn DEV PERPEN CENTERLINE=ON to show the correct position & ignore the form error of the sample plane.

Hope this helps.

[Abanks]

Use more than 3 hits, it's not enough, dead horse but 11 or 15 etc. Change the calculation to MAX_INSCR. If you use sample hits turn DEV PERPEN CENTERLINE=ON.

Hope this helps.

thank you. i will try anything. i just found where the MAX_INSCR and others are in the help menu. reading the info now. i didnt know there were options like that. thanks again

[JEFMAN]

You should also check the angle between the surface and the probe axis. If the hole is not // to CMM Zaxis ( A0B0 for example), the hits won't be in the same plane, so you will have an error... When you measure with calipers, you seach the best axis, you have to do the same here...

[Abanks]

ok so far i have changed the number of hits to about 15 per hole and slot. also changed to perp to centerline, and set to max inscr. so far that has not fixed the problem. not sure what else to do.

[tking]

John and Clayt have it, use a spherical tip and probe an autocircle with 3 sample hits (choose the depth to hit the shear not the break) and your issues will resolve.

.02,

TK

[Jimmy Hirsch]

with calipers

uh ............. got any gage pins ?

[Abanks]

ok thanks to all of you. I ended up changing to a speherical probe and now all my dimensions are in spec. Thanks again.

[JEFMAN]

Shank probe should be used with care...